CA2010828A1 - Chemical fixation process of liquid aqueous mud by implementation of a pouzzolanic reaction - Google Patents

Abstract

A liquid aqueous slurry is transferred into a cylindrical-conical tank (1) and it is agitated by blowing air with a plunging rod (2). A closed sludge circulation is established between the tank (1) and the reagent tank (3) using a recycling pump (4). There is an overflow. from the tank (3) to the tank (1). Add the lime in the tank (3) and the pozzolanic material in the tank (1). When the mixture of diluted sludge, lime, pozzolanic material is homogeneous, it is sent to a filter press (not shown) using a pump (5). The filter cakes are collected for final storage or for discharge for slow setting.

Description

2 ~ 1 ~ 82 ~

"Process for the chemical fixing of liquid aqueous mud by placing of a pozzolanic reaction "

The invention relates to a method of fixing aqueous mud by pozzolanic reaction.
When landfilled, residual sludge containing toxic products in their solid phase pose a risk to the environment because this solid phase is very finely dispersed and present therefore a very important interface with the external environment, which can only greatly favor the passage into solution toxic polluting species due to physical dissolution, hydrolysis, biodegradation, etc. . By elsewhere this sludge has an aqueous liquid phase interstitial itself more or less contaminated which constitutes the most mobile fraction.
The problem of environmental pollution is remedied by fixing this sludge using various so-called sludge fixation. Among these are the processes using pozzolanic reactions.
We define the pozzolanic power of a material mineral as the ability to fix lime to ordinary temperature to give a hydraulic binder, that is to say a compound capable of making taken in the presence of water ~ and therefore harden and solidify.
This effect is very slow and starts after 2 or 3 days after completion of the mixture The evolution of the material at ~ time is very slow and it may take several years to wait for maximum mechanical strength.
The most common pozzolanic materials are fly ash (dimension of the order of thirty microns). We can distinguish silco-aluminous ash from the combustion of hard coal and ash sulfo-calcics resulting from the combustion of lignite.
In practice, the implementation of the effect 35 ~ pozzolanlque fly ash can be implemented in , .. ... .
. .

- ~,, ~,,::
, - -. .::

. . .
,, 2 ~ ~ 82 ~

simultaneously adding these ashes and lime to the mud to treat.
Lime can also already be contained (in the form Ca (OH) 2 in the sludge (neutralization sludge or precipitation containing excess lime).
Commercial processes using pozzolanic reactions are already known from general works as RB POJASEK "Toxic and hazardous waste disposal" volume 1. Ann Arbor Science 1979 and USEPA, "Survey of solidification / stabilization technology for hazardous industrial wastes "EPA- 600 / 2-79-056 July 1979.
The POZ-O-TEC process developed by the Company American IUCS uses fly ash and other additives. Two reactions occur simultaneously: one lS rapid reaction between soluble salts on the one hand and lime and alumina present in the fly ash and a much slower pozzolanic reaction between silica and lime.
The immediately treated mud is physically present ~
in the form of a viscous pumpable fluid which will be left harden for several weeks at its storage site.
This process was developed to treat sludge from SO washing water from power station fumes then it was used to solidify waste liquids for the chemical and surface treatment industries metal.
In the Research-Cottrell process (see ref. RB
~ POJASEK above) variants of this process are used. The processing chain includes a gravity thickener, a centrifuge to dewater the mud, a conveyor fly ash, ~ a lime conveyor, and a mud mixer dehydrated / lime fly ash is lime.
In the case of wastewater treatment sludge metal surfaces a centrifuge or a press belt gives a residue which still contains about 80 ~ O by weight of water. A filter press makes it possible to obtain residues containing still ~ approx 70 ~ water.

:. .
~. . ,. . ,,. . :.

. ~
, - '''',.

2 ~

This high water content has drawbacks because high amounts of lime and pozzolan to get solid and this leads to large masses and volumes of solid waste to be stored.
In addition, the Research-Cottrell Process mixer must be powerful and robust because its function is to mix intimately a thick medium formed by the centered sludge after centrifugation with fly ash and lime.
The object of the invention is to propose a method of treatment of sparsely concentrated bees (for example around 3 to ~ by weight of dry matter per weight of sludge), more simple, because not using a powerful mixer, everything by obtaining a better separation between the solid phase of the mud and its liquid phase, i.e. recovering 35 to 40 % of the aqueous phase of the initial sludge. This aqueous phase can be recycled or subjected to a purification treatment in particular by ion exchange resins.
The object is achieved by the process according to the invention which consists of a chemical mud fixing process aqueous liquid, by carrying out a reaction pozzolanic, characterized in that the mixture is first of all liquid aqueous mud with pozzolanic material and possibly lime 5i the aqueous mud has not initially a sufficient content of calcium hydroxide to react with all of the pozzolanic reagent, so as to obtain a substantially homogeneous mixture which is then mechanically dehydrates to obtain a solid cake slowly solidifying and intended for storage.
In more detail, we perform in the above process successively the following steps:
a) if necessary adjust the pH of the sludge to a value greater than 9.
b) the product obtained in step a ~ is mixed with the pozzolanic material and possibly lime so the homogeneous mixture thus obtained has a content of sludge dry matter, calcium hydroxide, material pozzolanic in hydroxide weight ratios : ~ '''''~''.:

~: .........................................:

~. . ::. : -. . .

2 ~ 2 ~

calcium / dry matter between 1 and 2 and material pozzolanic / calcium hydroxide about 2.
c) performing mechanical dehydration of this ~ -homogeneous mixture.
In the process according to the inventi ~ n, lime must hear either quicklime CaO or slaked lime Ca (OH) 2.
Mechanical dehydration is called elimination plus or less complete of the liquid phase of a sludge by application of a force or pressure field, for example by filtration or centrifugation.
The process according to the invention has numerous benefits :
Agitation is carried out on a dilute aqueous medium relatively fluid, allowing faster better homogenization of solids (waste, lime, pozzolanic material) in the aqueous phase than in the Research-Cottrell process.
The pozzolanic material gives the previous mixture better aptitude for dehydration, in particular making filter cakes less compressible and more permeable than those resulting from the filtration of mud alone.
The solid phase collected solidifies slowly in a kind of stone which has good resistance to erosion and leaching.
; Finally this process can be implemented without other modification of a treatment plant that a storage and dosing of pozzolanic material.
In addition to the foregoing, the invention includes other provisions which will emerge from the description which follows which refers to setting examples implementing the process of the present invention and in the drawing schematic annexed in which:
~ the figure is a schematic view of an installation for implementing the method according to the invention.
EXAMPLES OF ~ OUES TREATMENT TESTS

-: ', -.,, - ~
. ~ -.

2 ~ 82 ~
s 1- Definition of sludge by their composition a) First mud V1.
In order to obtain c ~ mparahles tests between them a first U1 sludge was synthesized. U1 mud is obtained by neutralization with slaked lime of a salt acid lutic-n metallic.
The composition of the solid and liquid phases of this mud is reported in Table 1.
Board 0 C ~ ct rl-tlq ~ ¦ ~ - oll ~ ¦ P ~ - Il <lUId ~ _ 9 ~ 8 Re ~ l ~ tlvlté (- ~ .c ~ 15 .. I.
Reaid ~ l lO-C I 303 ~ 50 ~ 7163 . I - ---.
Residue ~ SSO-C I 256 260 1 ~ 9 ~ 5 __ Sulf ~ te ~ (S04--) _ 150 _, Cl ~ lor ~ lre: l (Cl-) -j 2961 I
t ~ ltrate ~ (N03-) 1 - ~ 254 II I
ISodiwa I S10 ~ 307 20jC lclu ~ 91210 I Z299 _ ll Pota ~ I 127 1 6.75 C ~ lum I 62 1 0.1 I Chrome 1 1750 1 57 25 t ~ etawt ICulvre I 2 ~ 60 0.058 , j Rer 221 ~ 0 0.133 - T
l ~ qean ~ sc ~ 9 ~ 1 '3.03 ~

.. _. _ ckrl 1360 d 0.22 P! Omb ~ 239 1 0 35 30: ~ 23 00 t 0.18 ~ 5 1 - -results expressed in mg / kg of mud results ~ expressed in mg / l of mud , ~. - '''~: -''';: ~. :, -, ,. .: ~: - -:

2 ~ 2 ~

This mud, with a dry matter content of approximately 261 g per kg of mud ~ 2fi ~ of dry solid matter), was diluted to 6 ~ by adding water to give the mud V1.
c) of ~ xth mud V2.
It is a waste sludge from industry treatment of metallic surfaces (chrome plating and nickel plating parts for the production of cycles) from a effluent treatment successively comprising:
- a chemical reduction of chromates by addition of sodium bisulfite in acid medium, - neutralization with quicklime up to a neighboring pH
10, at a rate of 14.1 g of lime / kg of effluent.
The essential characteristics of this mud are the following :
lS pH of the suspension liquid: 9.8 dry residue at 110-C: 60.7 g / kg concentration of metallic elements in salts ionizable (mg per kg of dry matter):
20 Sodiu ~: 8518 Calcium: 6175 Trivalent chrome: 980 Iron: 192 Nickel: 1310 2- Nature of p ~ uzzolanic materials used.
Different pozzolanic materials have been selected.
These materials are designated below under the reference respective A, ~, C, D, E.

A is silica smoke marketed by the Company French SOFREM;

~ is a commercial thermal silica having the composition next :

2 ~ 2 ~

Chemical analyzes:

ZrO2 1.3 ~

Fe23 0.2% Humidity 0.22 at 100-C
Na20 0.2 -0 Fire Loss (excluding H20) o, 3%
CaO 0.04 -0 between 100 and 1000-C
C 0.07 ~ pH of an aqueous suspension at 4%: 4.4 K20 0.06 ~ O
so3 0.006%
SiO2 94 'o Physical properties :
Theoretical density 2.2 g / cm3 15 Bulk density (not packed) 0.24 g / cm3 Specific surface (BET) 14 m2 / g Dimensions 90 ~ less than 2tm 80% lower ~
10% less than 0.2 ym J ~ 50 = 0.55 ~ m This thermal silica consists essentially of microspheres of vitreous silica.

It is a pozzolan marketed by TUBAG

D is a pozzolan of natural origin from VOLVIC (France) E is a material ~ based on ~ e silica, occurring in the form significant deposits of waste from the glass industry obtained as a residue from the old process of smoothing ~ glass.
Chemical analyzes (weight ~ 0) 90% SiO2 at least.
1.5 50 Na20 in its hydrated form .: '-.

. .
.
:
-:

2 ~ 82 ~

0.02 ~ of co-carbonated organic matter The rest being Fe203 The particle size of this powder is less than 5 ~ m The use of this material E hitherto unused is of great economic interest compared to the use of pozzolanic materials marketed which are generally relatively expensive I. Examples of implementing the method according to the invention in the treatment of svnthetic mud v1 The balance is reported in each test to 1 kg of mud gross V1.
Tests 1 to 6 are control tests.
Tests 7 to 22 are tests according to the process according to the invention.
To 1 kg of mud v1, either 60 g or 120 g of quicklime CaO and a double quantity of the material pozzolanic A, B, C or E. Then the mixture is homogenized by stirring before being subjected to filtration under pressure of 196 or 392 kPa.
As reported in table 2 for each test we measures the mass of the filtrate, the mass of final waste and its volume.

Table 2 (see next page) 2 ~ 2 ~

Table 2 I rv`u ~ lltite dc ré ~ ctif uti ~ Pr ~ iio ~ d ~ 11 ~ 5- d- filt ~ t 1 ~ 55- dt Volul ~ e ~ e Ess i no! lisé5 ~ gl-a ~ e5J (kg / c ~ 2) (gr ~) I d ~ ch ~ t fine waste; I
I chau ~ vive I pou ~ zol ~ ne 1 (9- ~ 5) l _ , 1 0 1 0 2, 878 122 91 ll .. __ I _ 2 1 0 j 0, 4 - 890 II ~ o 82; 5 ~ -T ------------ T -------- - _ _ _. _ _ _ _ _ _ _

3 1120 1 0 2,; 790 1,330,301 '

4,160 1 0 2 89 ~ 1,166,153 I. I.
It20 1 0 4 810 3l0 28S
TI - ~ 4 '-920 1 ~
- - ---- T ------------ T -----------. ~ _ _ _.
i 7 1120 1 A: 240 2 500 860 712 . .. _ , ~ ~ 60 d A: 120 2 645 53S, 4 ~ 3 i.,. _ 9. 1 120 1 A: 2 ~ 0,. . . 520 840 606 10 1 60 1 A 120 - '''",' 4 '!"'"",, - '.' 680 Soo 414 .Tl ~ 1 120 1 8: 240, 2 '575 7ss s39 , .. I. . ... _ li 1 60 ¦ B: 120 2 745 435 298 13 1 120 1 9: 240 4 '.- S9S 765 S00 1 '''----I, 14, 1 6 ~ 1 0: 120 4 765 - 41S 279 : - - T ----------__ ~ ___ ___ .__, _____ ___________ ____ _ _________.
.'I 120 i C: 2 ~ .0 2 '690 670- .. ,, 4S2', __. _. _ 6''1 60 ~ C: 120 2 800 380 269 . _ 17'1 1 ~ 0 d C: 2 ~ 0 4 ~ 22 638 404 1 --- ~. .... ~
8 1 60 j C: 120 ~ 810 370 250 r l20 1 ~: 240 2, ~ as 695 ~ 81 1 60 '1 ~: 120. 2. ~ 7,. 405 '279 r ---- - -;
2 1 ~ 20 '~: 240. ~ ~ 0 660 ~ 3 1;

- II.
2 2 1 60 I f ~: 120 ~ 1 ', 79 ~' 3q0 260 _. . . _ _. _ _ ~ _ .

-: -, ': "~' ~: ' - -.
- - '', :

20 ~ 828 The results obtained in these tests show the effect favorable filtration pressure which decreases the mass of final solid waste and increases the volume of filtrate obtained, the mass of final waste decreasing in order of materials A, ~, E, C.
Other tests reported below show the favorable influence of pozzolanic materials added to V1 mud on the filterability of it.
The filterability of a mud can be characterized by the following parameters:

1 'Specific resistance to filtration: ~
2- The compressibility coefficient of the cake: S
3- The limit dryness of the cake: ~ l These parameters have been described in the publication of AFNOR T97-001 of November 1979 (pages 1 to 14) having for title: Experimental standard - Sludge testing - Determination characteristics linked to the ability to concentration.
:: d is defined as the specific resistance to `20 filtration of the filtered suspension. It is independent of the concentration of it but basically depends on the size, shape and degree of agglomeration of the particles so ~ constituent liquids of the filter cake.
; Table 3 shows the specific resistances ~ obtained by adding lime and p ~ uzzolanic materials to v1 mud compared to that of v1 mud alone.
From the examination of this table we see that the materials pozzolanics C and n improve filterability.

,::

~. ~. - '. ~

. :
'~:
. . ,::
,: ,, `` ',. ':' 2al ~ s2s Table 3 IP-cssionlPentt de l-lCo ~ fficicnt dtl ~ és; st ~ nc ~ svccifi ~ ue fo- ~ ul ~ ¦ ¦right ~ ¦ co ~ rcl ~ tion I at ~ ~ n 1 ~ 1 ~ 3Cg ~ tlY_f (v) I
60u ~. I OS - r 0.215 1 0.999 j I, 37 10 2 1 0.0895 d 0.999 1 2.29 ~ 08 co III '1 -'- -t 'I 0.0611 1 0.999 1 3.12.10 1 0.0-83 1 0.999 1 ~, l3 1o8 aoy ~ - I OS I $ 0.1 ~ 3 10,999 1 1.01.10 rj -.
I 2 1 0.0 ~ 15 1 '0.999 I t.a3 10a co I - ~ - ~ -I ---- - I a 1 0-0- ~ 8 1 0 ~ 9991 2. ~. lO
7 ¦ 0.032 ~ ¦ 0-959 1 2.9! 3. ~ Oa Aug ~. ~ jO.S j 0.0-52 1 0.999 I Z. I ~ .107 j 2 j 0.0197 I0.999 I 5.0 ~ .107 C ~ O, I - III
1 0 01 ~ 2 1o.sse 1 6.76.107 0.0108 10.999 1 9-59- 10 ~
-~ 30ue - I0.5 1 0, 121 1 0.999 1. ~ ~. 10 ~
j 2 1 0.0320 1 0.9 ~ 3 T 8.19 107 C ~ O. ,. r __ II. _ I 1 '. :) 2 ~ 6 10.99 ~ 1 1.053 108 OI. I _- I
1 7 1 0_0139 10.996 1 I.i '-. 10 60uc s ~ ul ~ j 0.51 0.0611 j 0.999 1 9 ~ 9 107 (tc ~ olnl II _.. I -'-- I
2 i 0-0299 10.999 1 1 .86. Io8 0.022- 10.999 1 2.78. ~ O8 ~
t 1 0.0191 10.99a 1 __ ~ ~, :
,.,, -. . . . . . . .
... ....,:., ,,,.,. ,. "~,",. ..
, ::::,.:,. - .. '~: -:
:.: ~:. ,::

-. :. . . :.
:,:., ',. '' ~ .- ~. -. - ,,:
. . -,. ,:. -, -. .
:,. ~,. ,,: ~
- '.' :,,. ~:. ~:

2 ~ g28 Table 4 shows that the addition of the reagents pozzolanics causes a very significant decrease in S.

Table 4 Formulation ¦ Cor-l factor Coefficient of I relationship I compressibility T ~
aoue l I
Ca O ~ 1 0.994 1 0.427 '' Mud ICaO- j! 0.998 1 0, ~ 03 I
I admit ~ I
I ca o, 1 0.998 ¦ 0.556 aoue ~ l j Ca O ~ 1 0.913 j 0.177 I ~ o ~ e 1 0.99 ~ 1 0.712 I

Table S shows the limit dryness obtained by additions of pozzolanic materials in comparison with mud V1 alone.

.
'''' ": ',:'' - ~

2 ~ 2 8 Table 5 I Pressure in kg / cm2 ~ ormule ¦
r 2 1 4 1 7 80ue + lll CaO + ¦ 44.5 ~ 1 49.5% 1 53%

~ oue + lll Ca O + I ~, 8% 1 51.9% 1 55.5%

Mud + ll I
CaO + 1 50.7 ~ 1 5a, 9% 1 57, ~%
III
Mud + I S4% I r DI 1 60 ~ 51 ~ 6 ~

80ue ~ eul ~ 33% ¦ 36.9% ¦ ao, 8%
~ m ~ in) I
The previous results recorded in ~ Tables 2 to 5 show that p ~ uzzolanic materials improve the filterability of synthetic mud v1.

Other tests reported below in table 6 show the solidification results cakes obtained by tests n-1 to 22 appearing in the table 2.

Procedure:

Immediately after filtration, samples are taken cylindrical cylinders (diameter 3.5 cm, height

5.5 cm) of the different filter cakes and leave the solidification take place for 28 days for a first. .
batch and 6 months for a second batch. The test pieces are put then in water without agitation and we observe their appearance .
. ,. '-''-': - -:, : -,.

'''' ~.

-2 ~ 1 ~ 828 up to 15 hours after their immersion.

The results are summarized in Table 6 below with the following code:

O: burst. immediate ~. or very short term material immersed in water;
immersed in water;
+: total disintegration of the material after 15 hours;
+ ~: partial disintegration of the material, obtained in the state final small fragments (<15 mm);
+++: partial disintegration, ~ bt.ent.ion in the final state of large fragments (> 15 mm);
++++: cracking of the material;
+++++: physical state unchanged ++++++ unbreakable manually.

Table 6 (see next page) .

2 ~ 2 8 Water table 6 Ess ~ inj Observations on Observations on 28 OJ ~ IF 6 months +++ f I +++

1 2 1 ++++ d +++
3! ++++! +++
.4 1 ++++ I +++
5 1 ++++ I _- ++

6 1 +++++ I +++

7 I ++++ I ~ ++++++

8. 1 ++++ I ++++

9., 1 ++++ I +++++
1 0 1 ++++ I +++
-11 '' I +++ I ++++
12 1. +++ I +++
13 1 ++ I ++++++
14 1 +++ I ++++
I
15 d +++++ ij ++++++
1 ~ I +++++ I ++++
_ 1 7 1 +++ ~ +

1 8 1 +++++ I +++++
1 ~ d +++++ I ++++++
~.
20 .1 +++++ 1, +++++
1 '21 1 +++++''I+' +++++
, i 22 d +++++ d +++++

-. , -. ~.

20 ~ 0 ~ 28 Table 6 shows that the behavior of the treated materials according to the pr ~ cede of the inventian is generally hien better than mud alone or mud conditioned only. with lime.
Other tests (leaching tests) have been shown in pollutant retention material, only after a ripening of 6 months, the following metals: Cr, Ni, Cu, Fe, Cd, Mn were no longer detected in the aqueous phases (leachate) obtained in contacting without stirring for 15 hours, 100 g of cake covered with 1 liter of water.

II Example ~ the implementation of the process according to the invention in the treatment of mud v2 _ Figure 1 shows a schematic view of a installation for implementing the pr ~ ced selan the invention.
We transfer the b ~ ue v2 to a tank cylindro-conical 1 of about 20 m ~ and. we shake it by blowing air through a diving stick 2. A
reagent 3 of approximately 300 l located upstream relative to ~: to tank 1 allows the addition of lime.
This tank 3 is supplied with baue from the tank : 1 by permanent pumping using the recycling pump 4.
This tank 3 overflows into the tank 1 and it is filled ~:: of the same mud v2, d ~ nt the mass is agitated by the current -: ~ ferry caused by p ~ mpe 4.
The process includes the following steps:
: aj on ~ add an excess of powdery quicklime in the reagent tank at the rate of 11 g of lime per kg of mud : raw .: After a few minutes the lime introduced will found fully mixed with everything.
~: present in the installation. The pH measured is 12.7.
b) is added directly into the cylindrical tank conical 1 the chosen pozzolanic material (thermal silica B
in this case) to. grapes of 11 g per kg of initial berry;
c) stirring is continued in tank 1 for 10 3s ~ to ~ 5 minutes. We mix the mixture to the filter press (nan : - -",.,. ~.

.

~ Q ~ 2 ~ `

shown) using pump 5;
d) the mixture is sent to a filter press trays ~ trays 60 x 60 cm) and presses are made discontinuous under 4 kg / cm c) the filter cakes are collected to be put in final deposit or landfill with compaction possible.
Samples are also. collected.
After 28 days the cylindrical samples (diameter 3.5 cm, height 5.5 cm) are very consistent hard and unbreakable manually. Leaching test practiced without agitation under the same conditions as for the mud v1 shows the advantages of the process according to the invention by compared to simple liming (Table 7).

Table 7:

mud v2: lixiavihilit.é of pollutants after treatment.
Solubilized quantities ~ es during a leaching, in mg per kg of treated material.

Table 7 (see next page) :

- -:, .
, .

:.
. '''' ~ ~ "~ '.'.

Table 7 ! \, T.aitement ~ undergone IC ~ aUlage I Treatment Species ~ ¦ ¦ (lime + pozzolan) chimic, ~ Ps consi ~ erées Total salinity ¦ 10,240 ¦ 7,880 .. I --1- - ---Sulfates ¦ ~ 390 ¦ 2,800 . _ Sodium 1 3 100 1 2 lO0 - ~ --- ~ I
Calcium 1,700 ¦ 310 C. irome tr ~ t2le ~ t ¦ 220 ¦ 80 F ~ r I ~ 1 ¦ ~ 1 ickel: I ~ 0.5 ¦ ~ 0.5 pH of I l ~ a ~ ¦ 11.6 ¦ 6.0 ~ ~:
::::

, . :,:. ,. ,, -::. `:.

, -:
:: ~
~: '. ::: '', ~ ': -. .
.

Claims (5)

1. Process for the chemical fixing of aqueous mud liquid, by carrying out a pozzolanic reaction, characterized in that the aqueous hoe is first mixed liquid with pozzolanic material and possibly lime if the aqueous mud does not initially have a content sufficient calcium hydroxide to react with the all of the pozzolanic reagent, so as to obtain a substantially homogeneous mixture which is then dehydrated mechanically to obtain a solid cake which mass slowly and intended for storage. 2. Method according to claim 1, characterized in what liquid aqueous slurry has a dry matter content between 3 and 20% by weight. 3. Method according to claim 2, characterized in the following steps are carried out successively:
a) if necessary adjust the pH of the sludge to a value greater than 9;
b) the product obtained in step a) is mixed with the pozzolanic material and possibly. lime so the homogeneous mixture thus obtained has a content of sludge dry matter, calcium hydroxide, material pozzolanic in hydroxide weight ratios calcium / dry matter between 1 and 2 and material pozzolanic / calcium hydroxide about 2.
c) the mechanical dehydration of this homogeneous mixture. 4. Method according to any one of the claims 1 to 3, characterized in that the pozzolanic material used is a residue from glass softening, based on silica. 5. Method according to claim 4, characterized in that the residue from the glass softening is a powder of a particle size less than 5µm and containing at least 90%
weight of silica.